Modern health care is complex, and that complexity is increasing due to advances in research, new technologies, increasing intricacies in interdisciplinary care delivery, and evolving regulatory and payment structures. Our ability to process massive amounts of information, however, has remained constant.
Cognitive science can offer useful frameworks for understanding how to support clinicians in processing an ever-growing body of information. Cognitive load theory can be helpful, as it outlines how humans filter, process, store, and retrieve information necessary for decision-making.
First, we gather new information through reading, listening to, or imitating others. Most of our formal education comprises this type of data acquisition. We store the knowledge in long-term memory. When we need this information to make clinical decisions, we retrieve it from long-term memory, process it through working memory, and bring it to the forefront of awareness for clinical care. Simultaneously, we process data from the environment through working memory to take appropriate actions.
A key insight from cognitive science is that both new information and existing knowledge that needs to be retrieved must pass through working memory. According to cognitive load theory, working memory is a limited resource when dealing with novel information, but unlimited when dealing with familiar information.
As a consequence, when an individual is overloaded with information or mired by distraction caused by unnecessary information, the ability to integrate new, important information or alter stored information becomes limited, with direct implications for our ability to learn and to conduct patient care.
Overview of Cognitive Load Theory
Cognitive load theory describes the factors that influence how much working memory is occupied. Intrinsic cognitive load is the cognitive “weight” of the information or task, determined by the complexity of the material being processed, and this is fixed. A more difficult case or a very complicated procedure has a higher intrinsic load than a more straightforward case or procedure.
Extraneous cognitive load is the mental load imposed by the organization of information or a task. For example, a poorly organized unit, where one must go to multiple locations to acquire the requisite materials for a procedure, imposes unnecessary extraneous load. When extraneous load increases, it steals limited working memory, reducing humans’ ability to attend to complex information.
Standardization and simplification reduce extraneous load. For example, central line kits decrease the extraneous load by gathering everything needed in one place. Similarly, an electronic health record (EHR) can display all of a patient’s diabetes-related data on one screen, decreasing the extraneous load of accessing the information and freeing working memory to process the clinical data.
Implications for Hospital Workflow Design
With increasing complexity comes greater cognitive exhaustion for frontline providers. This is due to two phenomena well described by cognitive load theory. The first, split-attention effect, occurs when clinicians must interact with multiple sources to acquire and synthesize information essential to complete a clinical encounter. Examples of this are when clinicians must access several electronic information systems, multiple parts of a single EHR, or several hospital staff members to collect necessary data or insights into a patient case.
The second contributor is the redundancy effect, when the same information is concurrently presented in multiple ways. Clinicians frequently receive notification of the identical information via a page, a separate call from the lab, a flag in the lab page, and an email — all in the name of patient safety. These well-intentioned notifications create further noise, or extraneous cognitive load, making it more difficult for clinicians to process other data. In addition, these alerts interrupt unrelated work processes — for example, a physician trying to enter orders for a sick patient may get distracted by an alert notifying her of a medication allergy she has already discussed with the patient and pharmacy staff.
Redundancy can be mitigated by establishing a single workflow for urgent notifications that is reliable and not duplicated across multiple service lines. The argument that patient safety necessitates these multiple notifications does not take into account variability in the reliability of each party contacting the physician with the information or the impact of these notifications on other critical data pieces the physician is trying to process. A single workflow would reduce redundancy and decrease the impact of the attention-splitting effect of these notifications.
Solutions for Cognitive Exhaustion
Designing electronic health records and health care systems with these cognitive principles in mind can avoid the risk of overloading clinicians. For example, an EHR that organizes relevant patient information for a disease in one window as described above reduces the consequences of split-attention effect. A physician would have to open only a single window to see a patient’s glucose measurements, laboratory results, medication doses, and recent related clinical visits. This reduces the redundancy of data presentation and eliminates split attention.
Medicine can borrow a page from the aviation industry to address interruptions, which split attention and can create redundancy. Standard aviation industry training recognizes that when we divert our attention from one topic to another, there is an attentional “blink” during which we do not attend to roughly 90 seconds of data. Teams receive training to triage the urgency of the interruption through a shared framework of the importance of particular types of interruptions. If the interruption is of high urgency, staff members are trained to note the task they are working on by saying out loud, for example, “step 6 of landing checklist,” proceed with the interruption, then return to the original task. By “bookmarking” the task, they are recruiting the entire teams’ attention to the task at hand.
In health care, this could work during inpatient rounding. If a member of the medical team has an urgent issue to discuss with the team during rounds, he or she could approach team members and tell them they need to interrupt their clinical decision-making. The team would then note their current activity, “discussing volume overload and recent dry weights,” for example, before addressing the team member and responding to the urgent need.
Implications for Medical Education
Medical students are responsible for mastering an ever-growing volume of information, so we must also develop educational strategies that decrease extraneous load. One solution is to avoid presenting information in ways that are overly complex, such as overloaded slides and gratuitous PowerPoint effects. We must also help students develop the skill of distinguishing the signal (intrinsic load) from the noise (extraneous load) by teaching them to think out loud as they evaluate data, asking themselves questions such as, “What data leads me to that conclusion? What data doesn’t fit with that conclusion? What are the five to seven most important pieces of data for this patient today?” This helps learners focus on the key information.
Another strategy is using approaches that optimize the intrinsic load for learners. While the complexity of information cannot be altered, complicated concepts can be spaced out over time. This means presenting simpler mental models — the mind’s “tools” that help with problem-solving and decision-making — to novice learners and adding nuance or complexity as the learner advances.
Educators can help learners build mental models by beginning with simplistic representations, even if they are slightly inaccurate, and slowly adding detail over time. This spacing allows for working memory fatigue to recover. For example, initially presenting the cardiac system as two separate pumps may help illuminate the conceptual underpinnings of the cardiovascular system, before moving on to its actual complexity. This enables students to stretch their mental model so they can grasp more complex information later. In contrast, presenting complex data to novices will drown them in more data than they have the framework to process.
Finally, emerging neuroscience is highlighting more efficient teaching strategies designed to help learners create their own mental models, such as visualization through graphic organizers and concept maps. These tools build visual representations of the connections among data that a more experienced physician could hold in her mind. For example, a team could map all data a patient presents with and group data with potential diagnoses to visually depict which data fit with each diagnosis and which data are outliers.
Implications for the Future
Health care redesign efforts must accommodate the limitations of human cognition by creating optimal environments for learning, simplified information presentation, and less complicated workflows that create routines and predictability. Without such interventions, many providers will become overwhelmed by the number of patients they see, the complexity presented by those patients, and the systems within which they deliver care. With thoughtful attention to cognitive load theory, opportunities can be created to enhance education, systems design, and effective practice to improve clinician performance.